Research On Radar Forward Looking Imaging Algorithm And Application In Power Line Detection Applied Research

Radar forward-looking super-resolution imaging refers to the super-resolution imaging of the -10°～+10° area directly in front of the aircraft by radar.It’s a difficult and hot point in radar imaging research field,and also one of the key problems to be solved in the realization of helicopter collision avoidance.The scattering cross-sectional area of the power line is very small,and it is difficult to be detected during helicopter operations,resulting in flight accidents.Limited by Doppler bandwidth and platform size,traditional high-resolution SAR and real aperture imaging are not suitable for forward-looking imaging.Therefore,the research on improving radar forward-looking super-resolution imaging is of great significance to the realization of helicopter detection of power lines.In this paper,aiming at real aperture scanning radar,focusing on the forward-looking super-resolution imaging technology based on deconvolution,from the perspective of signal processing,the problems existing in the algorithm are deeply studied and analyzed.In order to solve the super-resolution problem of power lines in actual project scenarios,two new super-resolution imaging algorithms are proposed.The main research contents of this paper are as follows:(1)Under the scanning radar forward-looking imaging system,according to the geometric relationship between the moving platform and the target and the analysis of the echo characteristics,the radar forward-looking imaging model and the echo azimuth convolution model are established,which lays a theoretical foundation for subsequent research.(2)Aiming at the problem of poor resolution due to the lack of target prior information in radar echoes,Renyi entropy is introduced as prior information,and by adjusting the adjustable parameters in Renyi entropy,the proposed super-resolution algorithm can be more accurate.It is well suited for real target scenarios and improves the adaptability of the algorithm.(3)For different ground and sea scenes,under the Bayesian framework,super-resolution algorithms based on Gaussian noise distribution and Rayleigh noise distribution are proposed respectively.At the same time,entropy is introduced as a priori information to improve the noise suppression ability of the algorithm.A more targeted super-resolution imaging is achieved.(4)Aiming at the defects of the traditional super-resolution method in the recovery of target details,the regularization theory is used to add a weight matrix to the loss function,and the directional total variation operator and entropy are introduced as constraints,which significantly improves the performance of the super-resolution algorithm to targets in the low SNR environment.